1 paper · 1 filter
Hongfu Gao, Feipeng Zhang, Wenyu Jiang +3
Large language models (LLMs) have shown impressive performance on downstream tasks by in-context learning (ICL), which heavily relies on the quality of demonstrations selected from…